Landsat-Based Land Cover Classification for Cold Lake Oil Sands Region

A land classification raster dataset derived from 2010 Landsat imagery for the Cold Lake oil sands area. It contains 13 distinct land cover classes, including water bodies, forests, developed areas, and fire scars. The dataset was produced by the Government of Alberta as part of the Lower Athabasca Regional Plan implementation.

Use Cases

• Train a semantic segmentation model to classify land cover types like 'coniferous forest', 'water bodies', and 'developed areas' from satellite imagery.
• Analyze the spatial distribution of 'fire scar' and 'transitional bare surfaces' to assess environmental disturbance and recovery.
• Monitor changes in 'agriculture areas' and 'shrub land' over time to understand land-use conversion pressures near oil sands development.
• Use the 'mixed developed areas' and 'developed areas' classes to quantify urban and industrial footprint within the regional plan boundary.

Strengths

• Contains 13 distinct, defined land cover classes for model training and analysis.
• Covers a specific, high-interest region (Townships 56 to 69, Ranges 1 to 11) relevant to oil sands monitoring.
• Derived from a consistent, wall-to-wall source (2010 Landsat imagery).

Limitations

• Single temporal snapshot (2010) limits analysis of land cover change over time.
• The 'unclassified' class (0) may represent areas of poor data quality or ambiguity.
• Spatial resolution is limited by the source Landsat imagery (~30 meters).

Provenance

Source

Government of Alberta

Collection Method

Derived from 2010 Landsat satellite imagery.

Time Range

2010

Freshness

null

Geography

Cold Lake oil sands area, Alberta, Canada (Townships 56 to 69, Ranges 1 to 11, west of the 4th Meridian).